Skin Cancer Detection using Digital Image Processing and Implementation using ANN and ABCD Features

International Journal of Trend in
International Open Access Journal
ISSN No: 2456
@ IJTSRD | Available Online @ www.ijtsrd.com
Skin Cancer Detection
Implementation
Khaing Thazin Oo1
1,2
Department of Electronics Engineering,
Pyay Technological University, Myanmar
ABSTRACT
Melanoma is a serious type of skin cancer. It starts in
skin cells called melanocytes. There are 3 main types
of skin cancer, Melanoma, Basal and Squamous cell
carcinoma. Melanoma is more likely to spread to
other parts of the body. Early detection of malignant
melanoma in dermoscopy images is very important
and critical, since its detection in the early stage can
be helpful to cure it. Computer Aided Diagnosis
systems can be very helpful to facilitate the early
detection of cancers for dermatologists. Image
processing is a commonly used method for skin
cancer detection from the appearance of affected area
on the skin. In this work, a computerised method has
been developed to make use of Neural Networks in
the field of medical image processing. The ultimate
aim of this paper is to implement cost
emergency support systems; to process the medical
images. It is more advantageous to patients. The
dermoscopy image of suspect area of skin cancer is
taken and it goes under various pre
technique for noise removal and image enhancement.
Then the image is undergone to segmentation using
Thresholding method. Some features of image have to
be extracted using ABCD rules. In this
Asymmetry index and Geometric features are
extracted from the segmented image. These features
are given as the input to classifier. Artificial Neural
Network (ANN) with feed forward architecture is
used for classification purpose. It classifies the
image into cancerous or non-cancerous. The proposed
algorithm has been tested on the ISIC (International
Skin Imaging Collaboration) 2017 training and test
datasets. The ground truth data of each image is
available as well, so performance of this wor
evaluate quantitatively.
International Journal of Trend in Scientific Research and Development (IJTSRD)
International Open Access Journal | www.ijtsrd.com
ISSN No: 2456 - 6470 | Volume - 2 | Issue – 6 | Sep
www.ijtsrd.com | Volume – 2 | Issue – 6 | Sep-Oct 2018
Skin Cancer Detection using Digital Image Processing
Implementation using ANN and ABCD Features
1
, Dr. Moe Mon Myint2
, Dr. Khin Thuzar Win
1
Assistant Lecturer, 2,3
Professor
Department of Electronics Engineering, 3
Department of Mechronics Engineering
Pyay Technological University, Myanmar
Melanoma is a serious type of skin cancer. It starts in
skin cells called melanocytes. There are 3 main types
of skin cancer, Melanoma, Basal and Squamous cell
is more likely to spread to
other parts of the body. Early detection of malignant
copy images is very important
and critical, since its detection in the early stage can
be helpful to cure it. Computer Aided Diagnosis
elpful to facilitate the early
detection of cancers for dermatologists. Image
processing is a commonly used method for skin
cancer detection from the appearance of affected area
on the skin. In this work, a computerised method has
e of Neural Networks in
the field of medical image processing. The ultimate
aim of this paper is to implement cost-effective
to process the medical
It is more advantageous to patients. The
of skin cancer is
taken and it goes under various pre-processing
technique for noise removal and image enhancement.
Then the image is undergone to segmentation using
sholding method. Some features of image have to
be extracted using ABCD rules. In this work,
Asymmetry index and Geometric features are
extracted from the segmented image. These features
are given as the input to classifier. Artificial Neural
Network (ANN) with feed forward architecture is
used for classification purpose. It classifies the given
cancerous. The proposed
algorithm has been tested on the ISIC (International
Skin Imaging Collaboration) 2017 training and test
datasets. The ground truth data of each image is
available as well, so performance of this work can
Keyword: Skin cancer, Segmentation, Feature
Extraction, Classification, Melanoma
I. INTRODUCTION
Skin cancers can be classified into melanoma and
non-melanoma. Melanoma is a malignancy of the
cells which gives the skin its colo
and it can invade nearby tissues. Moreover, it spreads
through the whole human body and it might cause to
patient death and non-melanoma which is rarely
spread to other parts of the human body. Malignant
melanoma is the most aggressive type
cancers and its incidence has been rapidly increasing
[1] [2] [3] [4]. Nevertheless, it is also the most
treatable type of skin cancer if detected or diagnosed
at an early stage [5]. The diagnosis of melanoma in
early stage is a challenging and fundamental task for
dermatologists since some other skin lesions may
have similar physical characteristics. Dermos
considered as the widely common technique used to
perform an in-vivo observation of pigmented skin
lesions [6]. In early detection of malignant melanoma,
dermoscopic images have great potential, but their
interpretation is time consuming and subjective, even
for trained dermatologists. Therefore, the need to
build a system which can assist dermatologists to get
right decision for their diagnosis has become very
important.
Image processing is one of the widely used methods
for skin cancer detection. Dermoscopy could be a
non-invasive examination technique supported the
cause of incident light beam and oil immersion
technique to form potential the visual investigation of
surface structures of the skin. The detection of
melanoma using dermoscopy is higher than individual
observation based detection [
Research and Development (IJTSRD)
www.ijtsrd.com
6 | Sep – Oct 2018
Oct 2018 Page: 962
Digital Image Processing and
ABCD Features
Dr. Khin Thuzar Win3
Department of Mechronics Engineering,
n cancer, Segmentation, Feature
Classification, Melanoma
Skin cancers can be classified into melanoma and
melanoma. Melanoma is a malignancy of the
cells which gives the skin its colour (melanocytes)
and it can invade nearby tissues. Moreover, it spreads
through the whole human body and it might cause to
melanoma which is rarely
spread to other parts of the human body. Malignant
melanoma is the most aggressive type of human skin
cancers and its incidence has been rapidly increasing
[1] [2] [3] [4]. Nevertheless, it is also the most
treatable type of skin cancer if detected or diagnosed
at an early stage [5]. The diagnosis of melanoma in
and fundamental task for
dermatologists since some other skin lesions may
have similar physical characteristics. Dermoscopy is
considered as the widely common technique used to
vivo observation of pigmented skin
n of malignant melanoma,
dermoscopic images have great potential, but their
interpretation is time consuming and subjective, even
for trained dermatologists. Therefore, the need to
build a system which can assist dermatologists to get
eir diagnosis has become very
Image processing is one of the widely used methods
for skin cancer detection. Dermoscopy could be a
invasive examination technique supported the
cause of incident light beam and oil immersion
otential the visual investigation of
surface structures of the skin. The detection of
melanoma using dermoscopy is higher than individual
detection [3], but its diagnostic

International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456
accuracy depends on the factor of training the
dermatologist.
The diagnosis of melanoma from melanocytic nevi is
not clear and easy to identify, especially in the early
stage. Thus, automatic diagnosis tool is more effective
and essential part of physicians. Even when the
dermoscopy for diagnosis is done with the expert
dermatologists, the accuracy of melanoma diagnosis
is not more than 75-84% [4]. The computer aided
diagnostics is more useful to increase the diagnosis
accuracy as well as the speed [5].
The computer is not more inventive than human but
probably it may be able to extract some information,
like colour variation, asymmetry, texture features,
more accurately that may not be readily observed by
naked human eyes [5]. There have been many
proposed systems and algorithms such as the s
point checklist, ABCD rule, and the Menzies
[2, 3] to improve the diagnostics of the melanoma
skin cancer.
The key steps in a computer-aided diagnosis of
melanoma skin cancer are image acquisition of a skin
lesion, segmentation of the skin lesion from skin
region, extraction of geometric features of the lesion
blob and feature classification. Segmentation or
border detection is the course of action of separating
the skin lesion of melanoma from the circumferential
skin to form the area of interest. Feature extraction is
done to extract the geometric features which are
accountable for increasing the accuracy;
corresponding to those visually detected by
dermatologists, that meticulously characterizes a
melanoma lesion.
The feature extraction methodology of many
computerised melanoma detection systems has been
largely depending on the conventional clinical
diagnostic algorithm of ABCD-rule of dermoscopy
due to its effectiveness and simplicity of
implementation [7]. The effectiveness of methodology
stems from the fact that it incorporates the classic
features of a melanoma lesion such as asymmetry,
border irregularity, colour and diameter (or
differential structures), where surveyable measures
can be computed.
Dermoscopy is a diagnostic technique that is used
worldwide in the recognition and interpretation of
copious skin lesions [4]. Other than dermoscopy, a
accuracy depends on the factor of training the
iagnosis of melanoma from melanocytic nevi is
not clear and easy to identify, especially in the early
stage. Thus, automatic diagnosis tool is more effective
and essential part of physicians. Even when the
dermoscopy for diagnosis is done with the expert
ermatologists, the accuracy of melanoma diagnosis
4]. The computer aided
diagnostics is more useful to increase the diagnosis
The computer is not more inventive than human but
le to extract some information,
like colour variation, asymmetry, texture features,
more accurately that may not be readily observed by
5]. There have been many
proposed systems and algorithms such as the seven-
and the Menzies method
] to improve the diagnostics of the melanoma
aided diagnosis of
melanoma skin cancer are image acquisition of a skin
lesion, segmentation of the skin lesion from skin
geometric features of the lesion
blob and feature classification. Segmentation or
border detection is the course of action of separating
the skin lesion of melanoma from the circumferential
skin to form the area of interest. Feature extraction is
xtract the geometric features which are
accountable for increasing the accuracy;
ally detected by
that meticulously characterizes a
The feature extraction methodology of many
detection systems has been
largely depending on the conventional clinical
rule of dermoscopy
simplicity of
]. The effectiveness of methodology
tes the classic
features of a melanoma lesion such as asymmetry,
border irregularity, colour and diameter (or
differential structures), where surveyable measures
Dermoscopy is a diagnostic technique that is used
n and interpretation of
copious skin lesions [4]. Other than dermoscopy, a
computerised melanoma detection using Artificial
Neural Network classification has been adapted which
is efficient than the conventional one and Melanoma
detection using Artificial Neural Network is a more
effective method compared to other.
II. Methodology
The following steps are implemented for classification
of skin cancer.
Image Acquisition
Pre-processing
Segmentation
Feature Extraction
Classification
The first step is the capture
image is acquired from the ISI
through MATLAB. The second step is image pre
processing, the pre-processing technique can be
applied to eliminate the irrelevant data contains in the
image. The skin lesion regions such as
oils, hairs are removed from the original image using
median filter techniques. The third step is image
segmentation: the goal of image segmentation is to
make simpler change the representation of an image
into something that is more meaning
analyze. The next step is image enhancement; to
improve the quality of the image so that the
consequential image is better than the original image.
In this step, the required region is segmented out and
detected the edge of the skin lesio
feature extraction and the final step is classification of
the skin lesion image.
Fig. 1System Block Diagram
A. Pre-processing
In this work, few previous processing techniques are
needed. Prior to segmentation, all
processed in order to minimize undesirable features
that could affect the performance of the algorithm
such as reflections, presence of the hair and colo
differences between images. The image is also
normalized to a unique size and shape.
normalization allows for the comparison of the region
features such as positions and sizes between different
images.
International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
Oct 2018 Page: 963
computerised melanoma detection using Artificial
Neural Network classification has been adapted which
is efficient than the conventional one and Melanoma
Neural Network is a more
effective method compared to other.
The following steps are implemented for classification
The first step is the capture image the skin lesion
image is acquired from the ISIC 2018 database
through MATLAB. The second step is image pre-
processing technique can be
applied to eliminate the irrelevant data contains in the
image. The skin lesion regions such as label, marks,
oils, hairs are removed from the original image using
median filter techniques. The third step is image
segmentation: the goal of image segmentation is to
make simpler change the representation of an image
into something that is more meaningful and easier to
analyze. The next step is image enhancement; to
improve the quality of the image so that the
consequential image is better than the original image.
In this step, the required region is segmented out and
detected the edge of the skin lesion. The fourth step is
feature extraction and the final step is classification of
System Block Diagram image
In this work, few previous processing techniques are
needed. Prior to segmentation, all images are pre-
processed in order to minimize undesirable features
that could affect the performance of the algorithm
such as reflections, presence of the hair and colour
differences between images. The image is also
normalized to a unique size and shape. This
normalization allows for the comparison of the region
features such as positions and sizes between different

B. Skin Lesion Segmentation
Image segmentation is an essential process for most
image analysis subsequent tasks. Segmentation
divides an image into its constituent regions or
objects. The goal of segmentation is to make simpler
or change the representation of an image into
something that is more meaningful and easier to
analyse.
Image segmentation is the course of action of
segregating an image into multiple parts, which is
used to identify objects or other relevant information
in digital images. Background subtraction, also known
as blob detection, is an emerging technique in the
fields of image processing wherein an image’s
foreground is extracted for further processing.
Typically, an image’s regions of interest are objects in
its foreground. Segmentation methods can be
classified as thresholding, region based, Edge based
and clustering. In this work, thresholding is used
because of the computationally inexpensive and fast
and simple to implement.
C. Post-processing
Once binary image of skin lesion has been obtained,
the image then needs to be post processed in order to
resolve any problems there may be as follows:
To prevent edges which are dark in many of the
images, from remaining as part of the lesion mask;
To reduce any effects of disturbing artifacts as far
as possible;
To set soft edges, to ensure there are not too many
recesses or projections and that there is certain
convexity in the resulting mask
These problems can be reduces by using
morphological operation such as opening, closing and
filling process.
D. Feature Extraction
The foremost features of the Melanoma Skin Lesion
are its Asymmetric Index, Border features, Colo
Diameter. Hence, this system is proposed to extract
the (9) Geometric Features, (1) Asymmetry Index and
(1) diameter of the segmented skin lesion. These
features are adopted from the segmented image
containing only skin lesion, the image blob of the skin
lesion is analyzed to extract the 11 features
Asymmetry Features: Major and Minor Asymmetry
Indices:
Geometric Features:
Image segmentation is an essential process for most
image analysis subsequent tasks. Segmentation
image into its constituent regions or
objects. The goal of segmentation is to make simpler
or change the representation of an image into
something that is more meaningful and easier to
Image segmentation is the course of action of
image into multiple parts, which is
used to identify objects or other relevant information
in digital images. Background subtraction, also known
as blob detection, is an emerging technique in the
fields of image processing wherein an image’s
extracted for further processing.
Typically, an image’s regions of interest are objects in
its foreground. Segmentation methods can be
classified as thresholding, region based, Edge based
thresholding is used
omputationally inexpensive and fast
Once binary image of skin lesion has been obtained,
the image then needs to be post processed in order to
resolve any problems there may be as follows:
dark in many of the
images, from remaining as part of the lesion mask;
To reduce any effects of disturbing artifacts as far
To set soft edges, to ensure there are not too many
recesses or projections and that there is certain
These problems can be reduces by using
morphological operation such as opening, closing and
The foremost features of the Melanoma Skin Lesion
are its Asymmetric Index, Border features, Colour and
Diameter. Hence, this system is proposed to extract
the (9) Geometric Features, (1) Asymmetry Index and
(1) diameter of the segmented skin lesion. These
features are adopted from the segmented image
containing only skin lesion, the image blob of the skin
act the 11 features.
Asymmetry Features: Major and Minor Asymmetry
1. Area (A): Number of pixels of the lesion.
2. Perimeter (P): Number of pixels
detected boundary
3. Greatest Diameter (GD): The length of the line
which connects the two farthest
4. Shortest Diameter (SD): The length of the line
connecting the two closest boundary points and
passes across the lesion centroid.
5. Circularity Index (CRC): It explains the shape
uniformity.
6. Irregularity Index A (IrA):
7. Irregularity Index B (IrB):
8. Irregularity Index C (IrC):
9. Irregularity Index D (IrD):
Diameter: Diameter in pixels
E. Classification
The main issue of the classification task is to avoiding
over fitting caused by the sma
skin lesion in most dermatology datasets. In order to
solve this problem, the objective of the proposed
model is to firstly extract features from images and
secondly load those extracted representations on
ANN network to classify.
III. Performance Evaluation Parameter
The most common performance measures consider
the model’s ability to discern one class versus all
others. The class of interest is known as the positive
class, while all others are known as negative. The
relationship between positive class and negative class
predictions can be depicted as a 2
matrix that tabulates whether predictions fall into one
of four categories:
True positives (TP): These refer to the positive
tuples that were correctly label
classifier. It is assumed that TP is the number of
true positives.
True negatives (TN): These are the negative tuples
that were correctly labelled by the classifier. It is
assumed that TN is the number of true negatives.
False positive (FP): These are the ne
that were incorrectly labe
assumed that FP is the number of false positives.
False negative (FN): These are the positive tuples
that were mislabelled as negative. It is assumed
that FN is the number of false negatives.
Accuracy can be calculated by using this e
Accuracy ൌ
୒୳୫ୠୣ୰ ୭୤ ୡ୭୰୰ୣୡ୲
୘୭୲ୟ୪ ୬୳୫ୠୣ୰ ୭୤
Oct 2018 Page: 964
A): Number of pixels of the lesion.
Number of pixels along the
): The length of the line
connects the two farthest
Shortest Diameter (SD): The length of the line
connecting the two closest boundary points and
passes across the lesion centroid.
Circularity Index (CRC): It explains the shape
Irregularity Index A (IrA):
Irregularity Index D (IrD):
Diameter in pixels
The main issue of the classification task is to avoiding
over fitting caused by the small number of images of
skin lesion in most dermatology datasets. In order to
solve this problem, the objective of the proposed
model is to firstly extract features from images and
extracted representations on an
Performance Evaluation Parameter
The most common performance measures consider
the model’s ability to discern one class versus all
others. The class of interest is known as the positive
class, while all others are known as negative. The
en positive class and negative class
predictions can be depicted as a 2 ൈ 2 confusion
matrix that tabulates whether predictions fall into one
P): These refer to the positive
tuples that were correctly labelled by the
assifier. It is assumed that TP is the number of
True negatives (TN): These are the negative tuples
ed by the classifier. It is
assumed that TN is the number of true negatives.
False positive (FP): These are the negative tuples
that were incorrectly labelled as positive. It is
assumed that FP is the number of false positives.
False negative (FN): These are the positive tuples
led as negative. It is assumed
that FN is the number of false negatives.
racy can be calculated by using this equation:
ୡ୭୰୰ୣୡ୲ ୮୰ୣୢ୧ୡ୲୧୭୬ୱ
୭୤ ୮୰ୣୢ୧ୡ୲୧୭୬ୱ
(1)

ൌ
TP ൅ TN
TP ൅ TN ൅ FP ൅ FN
As indicated in Equation (1), accuracy only measures
the number of correct predictions of the classifier and
ignores the number of incorrect predictions.
Sensitivity, also known as recall, is computed as the
fraction of true positives that are correctly identified.
FNTP
TP
ySensitivit
+
=
Precision, which is computed as the fraction of
retrieved instances that are relevant.
FPTP
TP
Precision
+
=
Specificity, computed as the fraction of true negatives
that are correctly identified.
FPTN
TN
ySpecificit
+
=
IV. Test And Results
A. Results of Segmentation Process
There are three main phases: namely pre
segmentation and post-processing in segmentation
part. The input of the system is dermoscopic image o
skin lesion as shown in Fig. 2. After resizing the input
image, this image will convert to the gray scale
in order to get grater separability between the lesion
and background healthy skin. The resultant
image has been displayed. Most of the dermoscopic
images have some artifacts such as oil, bubble hair,
noise etc. these artifacts are removed by
median filter on gray scale image. After this, the
Otsu’s thresholding method has been applied on the
gray scale intensity image. This gives the desired
segmented image.
After these steps morphological operations are used to
enhance the segmented image. In this research work,
morphological opening, dilation, erosion and closing
operations are used. A morphological closing of the
mask is performed using a disk of radius 500 pixels
followed by a dilation using a disk of radius 40 pixels.
These operations smooth the border of the mask. The
experimental results for this post-processing phase are
shown in Fig. 3.
, accuracy only measures
the number of correct predictions of the classifier and
ignores the number of incorrect predictions.
, is computed as the
s that are correctly identified.
(2)
which is computed as the fraction of
(3)
, computed as the fraction of true negatives
(4)
namely pre-processing,
processing in segmentation
part. The input of the system is dermoscopic image of
. After resizing the input
gray scale image
in order to get grater separability between the lesion
and background healthy skin. The resultant gray scale
image has been displayed. Most of the dermoscopic
images have some artifacts such as oil, bubble hair,
ved by applying
scale image. After this, the
Otsu’s thresholding method has been applied on the
scale intensity image. This gives the desired
After these steps morphological operations are used to
ented image. In this research work,
morphological opening, dilation, erosion and closing
operations are used. A morphological closing of the
med using a disk of radius 500 pixels
followed by a dilation using a disk of radius 40 pixels.
operations smooth the border of the mask. The
essing phase are
Fig.2. Testing result of pre
segmentation
Fig.3. Testing result of segmented image after dilation
and erosion
B. Results of Classification
The first stage of the system is to select skin lesion
images. This can click original image menu item from
main window GUI of the proposed system.
After loading the input image, it is needed to segment
the lesion image by using O
Technique. The segmented image obtained from
Otsu's thresholding has the advantages of smaller
storage space, fast processing speed and ease in
manipulation, compared with gray
usually contains 256 levels.
In feature extraction, the standard features such as
Asymmetry Index, Area, Perimeter, Major Axis
Length, Minor Axis length, Circularity Index,
Irregularity Index and diameter are
segmented test image as shown in Fig.
standard features are very useful to classify the
melanoma skin cancer more accurately.
After extracting the features, these features are given
as input to the classifier. The classifier produces
whether the image is Melanoma or not. For the
Melanoma condition, the classifier o
normal skin (not melanoma) the output is 0. By
pressing the classify button, the classification process
is performed and the result is displayed as shown in
Fig. 7 and 8.
Oct 2018 Page: 965
result of pre-processing and
segmentation
result of segmented image after dilation
and erosion
The first stage of the system is to select skin lesion
This can click original image menu item from
main window GUI of the proposed system.
After loading the input image, it is needed to segment
the lesion image by using Otsu's Thresholding
Technique. The segmented image obtained from
Otsu's thresholding has the advantages of smaller
storage space, fast processing speed and ease in
n, compared with gray level image which
extraction, the standard features such as
Asymmetry Index, Area, Perimeter, Major Axis
Length, Minor Axis length, Circularity Index,
Irregularity Index and diameter are extracted from the
ted test image as shown in Fig.6. These
ery useful to classify the
melanoma skin cancer more accurately.
After extracting the features, these features are given
er. The classifier produces
whether the image is Melanoma or not. For the
Melanoma condition, the classifier output is 1 and for
normal skin (not melanoma) the output is 0. By
pressing the classify button, the classification process
ult is displayed as shown in

Fig.4. Loading Input Image
Fig.5. After Segmentation
Fig.6. Extracted Features from Testing Image
4. Loading Input Image
5. After Segmentation
Fig.6. Extracted Features from Testing Image
Fig.7. Classification Results of Melanoma
Fig.8. Classification Results of Non
C. Performance Evaluation
To evaluate performance in this system, 40
images from a test data set are tested. The accuracy of
skin lesion classification system is calculated. Table
1 describes the performance of the proposed system.
TABLE I Performance of The Proposed System
N
o
.
Im
age
Set
TP
F
P
T
N
F
N
Acc
urac
y
1
Tes
t
ing
Set
18 2 19 1
92.5
%
Oct 2018 Page: 966
7. Classification Results of Melanoma
8. Classification Results of Non-Melanoma
performance in this system, 40 unknown
images from a test data set are tested. The accuracy of
system is calculated. Table
describes the performance of the proposed system.
Performance of The Proposed System
Acc
urac
y
Sensi
tivity
Pre
cisi
on
Spe
cific
ity
92.5
%
94.7
%
90
%
90.4
7%

V. Conclusions
In this work, the system for melanoma skin cancer
detection system is developed by using MATLAB.
Image segmentation is the first step in early detection
of melanoma skin cancer. To analyze skin lesions, it
is necessary to accurately locate and isolate the
lesions. In this thesis work, the Otsu’s method is the
oldest and simplest one. It has shown the best
segmentation results among the three methods.
Feature extraction is considered as the most critical
state–of-the- art skin cancer screening system. In this
thesis, the feature extraction is based on ABCD
of dermatoscopy. Algorithms for extracting features
have been disused. All the features have been
calculated based on Otsu’s segmentation method.
Using the Neural Network classifier, melanoma skin
cancer diagnosis with a training accuracy of 100%
and testing accuracy of 93% is achieved.
Computational time is around 14 seconds fo
lesion classification. For illustration, a graphical user
interface has developed in order to facilitate the
diagnostic task for the dermatologists.
Acknowledgment
Firstly, the author would like to acknowledge
particular thanks to Union Minister of the Ministry of
Science and Education, for permitting to attend the
Master program at Pyay Technology University.
Much gratitude is owed to Dr. Nyaunt Soe, Rector,
Pyay Technological University, for his kind
permission to carry out this paper. The author i
deeply thankful to her supervisor, Dr. Moe Mon
Myint, Professor, Department of Electronic
Engineering, Pyay Technological University, fo
helpful and for providing guidelines. Moreover, the
author wishes to express special thanks to Dr. Khin
Thu Zar Win, Professor and Head, Department of
Mechatronic Engineering, Pyay Technological
University, Pyay for her kindness and suggestions.
Finally, I would like to thank my parents for
supporting to me.
In this work, the system for melanoma skin cancer
detection system is developed by using MATLAB.
step in early detection
of melanoma skin cancer. To analyze skin lesions, it
is necessary to accurately locate and isolate the
lesions. In this thesis work, the Otsu’s method is the
oldest and simplest one. It has shown the best
the three methods.
Feature extraction is considered as the most critical
art skin cancer screening system. In this
d on ABCD-rule
Algorithms for extracting features
the features have been
calculated based on Otsu’s segmentation method.
Using the Neural Network classifier, melanoma skin
cancer diagnosis with a training accuracy of 100%
and testing accuracy of 93% is achieved.
Computational time is around 14 seconds for each
lesion classification. For illustration, a graphical user
interface has developed in order to facilitate the
Firstly, the author would like to acknowledge
the Ministry of
Science and Education, for permitting to attend the
Master program at Pyay Technology University.
Much gratitude is owed to Dr. Nyaunt Soe, Rector,
Pyay Technological University, for his kind
permission to carry out this paper. The author is
r supervisor, Dr. Moe Mon
Professor, Department of Electronic
Engineering, Pyay Technological University, for her
guidelines. Moreover, the
author wishes to express special thanks to Dr. Khin
in, Professor and Head, Department of
Mechatronic Engineering, Pyay Technological
University, Pyay for her kindness and suggestions.
Finally, I would like to thank my parents for
References
1. Xu L, Jackowski M, Goshtasby A, Roseman,
“Segmentation of cancer images”, Image and
Vision Computing, Volume 17, Issue1, pg. 65
2. J Abdul Jaleel, Sibi Salim, Aswin.
“Computer Aided Detection 01 Skin Cancer”,
International Conference on Circuits, Power and
Computing Technologies, 2013
3. Arroyo J L G and Zapirain B G
Detection 01 Skin Cancer”,
Computers in biology and
biology and medicine 44 14157
4. Dr. S. Gopinathan, S.
“Feature Extraction through Image Processing
Techniques”, International Journal of Emerging
Trends & Technology in Computer Science
(IJETTCS), Volume 5, Issue 4, July
Web Site:www.ijettcs.org
5. Uzma Bano Ansari, “Skin Cancer Detection Using
Image Processing”, International Research Journal
of Engineering and Technology (IRJET), Volume:
04 | Apr-2017
6. Shivangi jain, Vandana jagtap, Nitin Pise,
“Computer Aided Melanoma Skin Cancer
Detection Using Image Processing”, Internal
Conference on Intelligent Computing
Communication & Convergence, Vol 48, pp 725
740, 2015, ISSN 1877
http://www.sciencedirect.com/science/article/pii/S
1877050915007188
7. M. Elbaum, “Computer
diagnosis”, Dermatologic clinics, vol. 20, pp. 735
747, 2002
Oct 2018 Page: 967
ckowski M, Goshtasby A, Roseman,
tation of cancer images”, Image and
Vision Computing, Volume 17, Issue1, pg. 65-74.
J Abdul Jaleel, Sibi Salim, Aswin. R. B,
“Computer Aided Detection 01 Skin Cancer”,
International Conference on Circuits, Power and
Computing Technologies, 2013
and Zapirain B G, “Computer Aided
Detection 01 Skin Cancer”, 2014, Computers in
Computers in biology and .medicine 44 14157
biology and medicine 44 14157
Nancy Arokia Rani,
“Feature Extraction through Image Processing
national Journal of Emerging
Trends & Technology in Computer Science
sue 4, July-August 2016,
Web Site:www.ijettcs.org
Uzma Bano Ansari, “Skin Cancer Detection Using
Image Processing”, International Research Journal
and Technology (IRJET), Volume:
Shivangi jain, Vandana jagtap, Nitin Pise,
“Computer Aided Melanoma Skin Cancer
Detection Using Image Processing”, Internal
Conference on Intelligent Computing
Communication & Convergence, Vol 48, pp 725-
015, ISSN 1877-0509,
.sciencedirect.com/science/article/pii/S
M. Elbaum, “Computer-aided melanoma
diagnosis”, Dermatologic clinics, vol. 20, pp. 735-

Skin Cancer Detection using Digital Image Processing and Implementation using ANN and ABCD Features

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to Skin Cancer Detection using Digital Image Processing and Implementation using ANN and ABCD Features

Similar to Skin Cancer Detection using Digital Image Processing and Implementation using ANN and ABCD Features (20)

More from ijtsrd

More from ijtsrd (20)

Recently uploaded

Recently uploaded (20)

Skin Cancer Detection using Digital Image Processing and Implementation using ANN and ABCD Features